1. Secure Unlocking of Mobile Touch Screen Devices by Simple Gestures – You can see it but you can not do it Muhammad Shahzad, Alex X. Liu Michigan State University East Lansing, Michigan, USA Arjmand Samuel Microsoft Research Redmond, WA, USA PRESENTED by Ivan Miguel Peñaranda Soria

2. Outline I.M.P.S Motivation Proposed Solution Technical Challenges Related Work Conclusion and Comments

3. Motivation I.M.P.S Password/PIN/Pattern

4. Proposed Solution I.M.P.S GEAT Gesture based authentication scheme for the secure unlocking of touch screen devices. 1° 15 - 25 times to obtain training samples 2° Extracts and selects behavior features from those sample gestures. 3° Builds models that can classify each gesture input as legitimate or illegitimate using machine learning techniques. How does it works?Adventage Is secure against smudge attacks whereas some biometrics. Does not require additional hardware for touch screen devices

5. Technical Challenges 1st. Challenge: Choose features that can model how a gesture is performed. I.M.P.S Velocity magnitude. Device acceleration. Stroke time. Inter–stroke time. Stroke displacement magnitude. Stroke displacement direction. Velocity direction.

6. Technical Challenges 2nd. Challenge: Segment each stroke into sub-strokes for a user so that the user has consistent and distinguishing behavior for the sub-strokes. I.M.P.S 3rd. Challenge: Learn multiple behaviors from the training samples of a gesture because people exhibit different behaviors when they perform the same gesture in different postures such as sitting and lying down. 4th. Challenge: Remove the high frequency noise in the time series of coordinate values of touch points. 5th. Challenge: Design effective gestures. 6th. Challenge: Identify gestures for a given user that result in low false positive and false negative rates.

7. Related Work I.M.P.S  Gesture Based Authentication on Phones  Phone Usage Based Authentication  Keystrokes Based Authentication  Gait Based Authentication

8. Experimental Results Results of evaluation of GEAT I.M.P.S Reports EERs from Matlab simulations on gestures in our data set. Study the impact of the number of training samples on the EER of GEAT Study the impact of the threshold of coefficient of variation on the EER of GEAT and justify our choice of using 0.1 as the threshold. Report the results from real world evaluation of GEAT implemented on Windows smart phones. Collected 15009 gesture samples from 50 volunteers.

9. Conclusion GEAT achieves an EER of 0.5% with 3 gestures using only 25 training samples. I.M.P.S Passwords/PINs/Patterns Vs. GEAT Identified seven types of features and this proposed algorithms to model multiple behaviors of a user in performing each gesture.

10. Comments I.M.P.S To make more than > 2 a point. Some ideas to make more secure GEAT Give some time to capture more than 1 stroke To use the signing of the users to unlock the phone

11. Thank you for your attention!